Blinding and exclusions after allocation in randomised controlled trials: survey of published parallel group trials in obstetrics and gynaecology

OBJECTIVE

To assess the methodological quality of approaches to blind ing and to handling of exclusions as reported in randomised trials from one medical specialty.

DESIGN

Survey of published, parallel group randomised controlled trials.

DATA SOURCES

A random sample of 110 reports in which allocation was described as randomised from 1990 and 1991 volumes of four journals of obstetrics and gynaecology.

MAIN OUTCOME MEASURES

The adequacy of the descriptions of double blinding and exclusions after randomisation.

RESULTS

Through 31 trials reported being double blind, about twice as many could have been. Of the 31 trials only eight (26%) provided information on the protection of the allocation schedule and only five (16%) provided some written assurance of successful implementation of double blinding. Of 38 trials in which the authors provided sufficient information for readers to infer that no exclusions after randomisation had occurred, six (16%) reported adequate allocation concealment and none stated that an intention to treat analysis had been performed. That compared with 14 (27%) and six (12%), respectively, for the 52 trials that reported exclusions.

CONCLUSIONS

Investigators could have double blinded more often. When they did double blind, they reported poorly and rarely evaluated it. Paradoxically, trials that reported exclusions seemed generally of a higher methodological standard than those that had no apparent exclusions. Exclusions from analysis may have been made in some of the trials in which no exclusions were reported. Editors and readers of reports of randomised trials should understand that flawed reporting of exclusions may often provide a misleading impression of the quality of the trial.

Sustaining interventions in community systems: on the relationship between researchers and communities

Important goals of research-based community interventions include the long-term maintenance of effects and fostering of collaboration between researchers and community leaders. This article reviews the challenges associated with transferring innovations to community systems, changing program delivery from an experimental context controlled by researchers to program delivery controlled by community organizations, and sustaining long-term effects of interventions. It is suggested that researchers who develop and implement community interventions in diverse health areas need to confront several issues: (a) fostering effective long-term relationships between researchers and the communities they study and in which they intervene and (b) designing and implementing interventions that are useful to community systems after the formal phase of research ends.

Effect of integration of injury control information into a high school physics course

STUDY OBJECTIVE

To investigate the effects on knowledge, attitudes, and self-reported behaviors of a 1-week course of injury control and crash safety information integrated within a high school physics curriculum.

METHODS

Students in an intervention high school (n=129) were compared with students in a control high school (n=74) enrolled in a comparable physics curriculum. A standardized survey was administered before instruction (time T1), and at 2 weeks (T2) and 6 months (T3) after instruction was completed. The behaviors measured were self-reported use of seat belts, speeding, drinking and driving, and intention to use seat belts in the future.

RESULTS

At T2, students in the intervention group reported attitudes that were less favorable toward risk-taking in regard to speeding and seat belt use than those of the control group. At T3, there was still a difference in attitudes toward speeding but not toward seat belt use. The intervention significantly altered the knowledge level of the course participants, and these changes persisted to T3. The strongest and most persistent change was that students in the intervention group reported increasing their use of seat belts when riding as a passenger. (Seat belt use as a driver was high for both groups.) The intervention group showed a significant increase in their 1-year intentions to use seat belts both as a driver and as a passenger.

CONCLUSION

This study demonstrated that driver safety education can be successfully integrated into a mainstream high school science curriculum. Future studies measuring the effects of this curriculum on observed behaviors are needed.

More informative abstracts revisited

Following proposals in 1987 and 1988, several medical journals have provided more informative abstracts ("structured abstracts") for articles of clinical interest. Structured abstracts for original studies require authors to systematically disclose the objective, basic research design, clinical setting, participants, interventions (if any), main outcome measurements, results, and conclusions; and for literature reviews the objective, data sources, methods of study selection, data extraction and synthesis, and conclusions. More informative abstracts of this kind can facilitate peer review before publication, assist clinical readers to find articles that are both scientifically sound and applicable to their practices, and allow more precise computerized literature searches. We review the feasibility, acceptability, and dissemination of structured abstracts, reassess the underlying strategy, and describe modifications of the approach. This innovation can aid communication from scientists to clinicians, and other clinical journals are invited to join this effort.

Randomised consent designs in cancer clinical trials

In 1977, Zelen proposed a new design for clinical trials with the aim of increasing recruitment by avoiding some of the problems associated with obtaining informed consent. These 'randomised consent' designs have proved controversial, and have not often been used. This paper explains the statistical aspects of single and double randomised consent designs and reviews some of the ethical issues. All identified published cancer treatment trials using a randomised consent design are considered in some detail. Reasons for and against the use of these designs are summarised.

Sustaining interventions in community systems: on the relationship between researchers and communities

Important goals of research-based community interventions include the long-term maintenance of effects and fostering of collaboration between researchers and community leaders. This article reviews the challenges associated with transferring innovations to community systems, changing program delivery from an experimental context controlled by researchers to program delivery controlled by community organizations, and sustaining long-term effects of interventions. It is suggested that researchers who develop and implement community interventions in diverse health areas need to confront several issues: (a) fostering effective long-term relationships between researchers and the communities they study and in which they intervene and (b) designing and implementing interventions that are useful to community systems after the formal phase of research ends.

Estimating sample sizes for binary, ordered categorical, and continuous outcomes in two group comparisons

Sample size calculations are now mandatory for many research protocols, but the ones useful in common situations are not all easily accessible. This paper outlines the ways of calculating sample sizes in two group studies for binary, ordered categorical, and continuous outcomes. Formulas and worked examples are given. Maximum power is usually achieved by having equal numbers in the two groups. However, this is not always possible and calculations for unequal group sizes are given.

Comparing methods of measurement: why plotting difference against standard method is misleading

When comparing a new method of measurement with a standard method, one of the things we want to know is whether the difference between the measurements by the two methods is related to the magnitude of the measurement. A plot of the difference against the standard measurement is sometimes suggested, but this will always appear to show a relation between difference and magnitude when there is none. A plot of the difference against the average of the standard and new measurements is unlikely to mislead in this way. We show this theoretically and by a practical example.

Aflatoxin and outcome from acute lower respiratory infection in children in The Philippines

Aflatoxin is immunosuppressive in experimental conditions. This study addressed its potentially contributory role in the poor outcome of acute lower respiratory infections (ALRI) in children in The Philippines. The catchment area included peri-urban slums and middle-class housing. One hundred and fifteen children (mean age 2.1, range 0.08-12 years) were enrolled and their serum and urine obtained at presentation with ALRI. Aflatoxins in serum and aflatoxin metabolites in urine were measured by previously validated ELISA tests. Using the 1986 WHO criteria for the severity of ALRI, 31% had mild, 12% moderate, 49% severe and 8% severe complicated pneumonia. Eighty of 97 (82%) chest radiographs were abnormal. Ninety per cent of the children were below average weight for age, using Filipino standards, with a mean of 79% (range 27-157%). Thirteen (11%) children died. Aflatoxin in their serum, reflecting recent ingestion, was detected in 33%, with a mean positive value of 462 pg/ml. Aflatoxin metabolites (reflecting chronic ingestion) were detected in 64 of 65 urines collected, with a mean value of 0.1-4.77ng/ml. None of the children with detectable serum aflatoxin died. Anorexia and impaired consciousness were strongly associated with a poor outcome (prolonged fever or death). There was a strong association between undetectable serum aflatoxin concentrations and death (p = 0.004), perhaps reflecting anorexia. There was no relationship between the concentration of urinary aflatoxin metabolites and outcome. Serum was also obtained from 29 mothers on admission and none contained detectable aflatoxin. As virtually all the children had evidence of exposure to aflatoxin, a potentially immunosuppressive role in the context of pneumonia cannot be excluded.

The methodologic quality of randomization as assessed from reports of trials in specialist and general medical journals

OBJECTIVE

To assess the quality of randomization from reports of trials in a sample of specialist journals, and to compare those results with a similar assessment from a sample of general medical journals.

DESIGN

Evaluation of all 206 reports of parallel-group randomized trials published in the 1990 and 1991 volumes of four journals of obstetrics and gynecology and of 81 reports of trials published during 1987 in four general medical journals.

RESULTS

Of the reports published in the specialist and in the general medical journals, only 32% and 48%, respectively, reported having used an adequate method to generate random numbers; only 23% and 26%, respectively, contained information showing that steps had been taken to conceal assignment until the point of treatment allocation; and merely 9% and 15%, respectively, described adequate methods of both sequence generation and allocation concealment. In those reports of trials that had apparently used unrestricted randomization, the differences in sample sizes between treatment and control groups were much smaller than would be expected by chance, and that feature was more marked in the specialist journals. In reports of trials in which hypothesis tests had been used to compare baseline characteristics, only 2% of tests reported in specialist journals and 4% of tests reported in general journals were statistically significant, lower than the expected rate of 5%.

CONCLUSIONS

Generating unbiased comparison groups requires proper randomization, yet the reports in these specialist and general journals usually provided inadequate or unacceptable information. Additional analyses suggest that nonrandom manipulation of comparison groups and selective reporting of baseline comparisons may have occurred.

Review of survival analyses published in cancer journals

Survival analysis has found widespread applications in medicine in the last 10-15 years. However, there has been no published review of the use and presentation of survival analyses. We have carried out a systematic review of the research papers published between October and December 1991 in five clinical oncology journals. A total of 132 papers were reviewed. We looked at several aspects of study design, data handling, analysis and presentation of the results. We found that almost half of the papers did not give any summary of length of follow-up; that in 62% of papers at least one end point was not clearly defined; and that both logrank and multivariate analyses were frequently reported at most only as P-values [63/84 (75%) and 22/47 (47%) respectively]. Furthermore, although many studies were small, uncertainty of the estimates was rarely indicated [in 13/84 (15%) logrank and 16/47 (34%) multivariate results]. The procedure for categorisation of continuous variables in logrank analyses was explained in only 8/49 (16%) papers. The quality of graphs was felt to be poor in 43/117 (37%) papers which included at least one survival curve. To address some of the presentational inadequacies found in this review we include new suggested guidelines for the presentation of survival analyses in medical journals. These would complement the statistical guidelines recommended by several clinical oncology journals.

Assessment of and intervention for psychosocial problems in routine oncology practice

An audit was carried out of 51 oncology patients referred to a clinical psychology service to identify the characteristics of patients selected for referral and to assess change following psychological intervention. A survey was conducted of an unselected sample of oncology patients representative of the workload of the oncology department from which the referrals came, to determine the prevalence of comparable psychosocial problems among patients who were not referred for help and to assess whether doctors were aware of the problems patients reported. Data were collected using the Hospital Anxiety and Depression (HAD) and Mental Adjustment to Cancer (MAC) scales and a problem checklist devised for this study. Referred patients were significantly more anxious and depressed (P < 0.001) and showed poorer adjustment on MAC scales than the surveyed sample, but 30% of the latter group warranted assessment for anxiety and 23% for depression. The number of psychosocial problems of their severity. Intervention was clinically significant mood disorder irrespective of the specific problems of their severity. Intervention was associated with a significant improvement in distress and problems for referred patients by the time of discharge. Psychosocial problems were often undetected by staff even in referred patients. The checklist is a feasible screening method for potentially remediable problems which are cumulatively a significant contributor to cancer patients' distress.

Repeatability of measurement of fetal nuchal translucency thickness

The aim of this prospective study was to assess the repeatability of measurement of fetal nuchal translucency thickness at 10-14 weeks' gestation. The nuchal translucency was measured by two of four operators in 200 pregnant women attending the Harris Birthright Research Centre for Fetal Medicine at 10-14 weeks' gestation. To assess repeatability of different components of variability, six measurements of nuchal translucency were made on each fetus, with a total of 1200 measurements. The data of this study demonstrate that 95% of the time the intraobserver, interobserver and caliper placement repeatability of measuring fetal nuchal translucency were less than 0.54 mm, 0.62 mm and 0.58 mm, respectively. In addition, the repeatability was unrelated to the size of the nuchal translucency. The findings of this study demonstrate that, when the nuchal translucency thickness is measured by well-trained operators, the measurement is highly reproducible.

Strategies for community health intervention: promises, paradoxes, pitfalls

This paper provides a rationale for intervening on shared determinants of health at the community level of analysis, presents strategies for intervening on shared determinant risks at the community-level of analysis, and illustrates through a case study of a youth program how a small-scale community intervention on shared determinants might work in practice. Although there is a long history of community intervention to improve health, high quality research on these interventions has been conducted only recently. In evaluating the usefulness of a community intervention, the distinction between relative risk and population-attributable risk is not always made clear. These two distinct types of risk are reviewed considering their implications for what risk factors are targeted by community interventions and what criteria are reasonable to apply in evaluating the utility of these interventions. A role for linking root cause analysis with attributable risk and interventions to reduce risk is presented.